A great deal has been written about the benefits of open access and it should be intuitively obvious to all librarians (and really anyone) that universal free open access will be the optimum for the benefit of information availability for humankind. Thus, the EU is moving to open access all papers they fund by 2020. How this will occur is still being determined. The problem with open access has always been how to fund the publishing of academic journals appropriately. This article propose a new method of funding universal open access to all academic fields using efficiency gains available to academics. Better still all of these efficiency gains are currently accessible to science funders and is in line with their missions.
Everyone's time is valuable, but science funders should pay particular attention to how academic researchers spend their time on funded projects. With the average IQ of a PhD in the top 5% of the population and the time to earn a PhD averaging 8 years society has made an enormous investment in their education. For this investment science funders should seek to have the highest return based on efficient use of their research time to generate new knowledge. Unfortunately, the current practices of funding, reporting, and disseminating scientific research suffers from three core inefficiencies that squander a shocking fraction of academic research time. Fortunately, there is a straight forward solution to significantly reduce wasted academic time from all of these losses: Science funders could support peer-reviewed open-access (OA) journals for their sub-disciplines. Remarkably these gains in academic efficiency come with the side benefit of providing universal open access to all future scientific literature.
First, in order to fund research, scientists must invest enormous quantities of their research time in grant writing. This is analogous to the waste associated with incumbent politicians fund-raising and campaigning rather than governing. Only a small fraction of this grant-writing time results in a successful funded project (e.g. NIH funding rates are about 21%, but even this value does not include the number of proposals that are disqualified for arbitrary reasons (e.g. formatting) that drops the real total funding rate for science to single digits in the U.S.). If we conservatively estimate that an academic must invest 2 weeks of research time to write a competitive proposal and wins only one per year, 10 weeks of productive research are lost per year, or about a fifth of her research time (19.2%). This is a huge loss for the average academic, which provides no new knowledge for society.
Second, once successfully funded, the faculty member has become under increased scrutiny for budget projections, spending rates, individual expense audits, compliance and reporting. In general, universities interpret oversight requests from funders, as mandates to hire internal auditors, which is partially responsible for adminstrative bloat and the concomitant overhead rate increases. This inefficiency is difficult to quantify as it varies by institution, but with overhead rates averaging over 53%, it is safe to assume that these direct losses from grant income for administration reduce productive research expenditures from funders like the U.S. National Science Foundation (NSF). Worse than this direct loss, faculty must spend precious research time responding to demands from compliance administrators for justifications for research expenditures. If we conservatively assume only one research day a year is wasted on this correspondence, another 0.4% of research time is lost internally. This loss, is trivial compared to external auditing and reporting. Older faculty can recall when reporting was done on a “per project” or an annual basis. Now quarterly reports are the norm and some government programs have moved to monthly reporting requirements. These reports are generally only read by a single project manager, who may not even be technically trained. If we assume a day per month is invested in reporting, another 4.6% of faculty research time is lost.
Lastly, the academic is expected to publish research in the peer-reviewed literature, which has enormous value for sharing science all over the world. It is clear that progress and the goals of science are best served by free and open access to the literature for all academics. Sadly, even this virtuous work is tainted by the increasing costs for access to the literature (e.g. Elsevier, the largest publisher, charges ~$900,000 annually for their holdings). With few institutions able to afford 1-click universal access to the whole of the peer-reviewed literature, research velocity is significantly hampered everywhere as all able minds are not working on the problems. In general, academics working without literature access are cursed to the digital equivalent of the post-card requests of yesteryear to peers for pre-prints. This trend again wastes academic research time everywhere servicing these requests. Academics who voluntarily post pre-prints on sites like arxiv.org and academia.edu can help reduce personal requests, but a reasonably active research group will lose at least another 16 hours (2 research days) per year making their work available (0.8% of faculty research time).
All together, these three inefficiencies cost average academics approximately 25% of their research time. Worse, these inefficiencies are increasing with time.
Let us consider the NSF, which could utilize their existing organization structure based around directorates and divisions to create a well-ordered offering of open access journals. The Division of Physics, would, for example, offer the NSF Journal of Physics, which could be further divided by sub-division/sub-discipline. To fully take advantage of the efficiencies possible by this process two new requirements will need to be implemented by funders.
First, proposals would be formatted in the form of the Introduction and Methods section of a journal article. In this way, the literature review and transparent methodology sections of an academic article would be accomplished at the proposal stage and simply have the results, discussion and conclusions sections filled in after the research was accomplished in the final article. This would eliminate redundant effort needed in the standard process. Special sections for individual funders (e.g. Intellectual Merit and Broader Impacts for the NSF) could be included and then used as sub-sections in the Discussion of the final article. Complicated grants would be sub-divided into several papers. Funding decisions would be based on the same process that is used currently for each funder and thus research quality would not be adversely effected. Similarly, all financial and compliance issues would occur as they do now without the necessity for any changes.
Second, to go beyond the NIH OA mandate, in place of reporting funded scientist would be required to publish their findings in the public domain in these free OA Journals. Then the reporting would only be minimized to collecting and reporting results and discussing their meaning for the entire world rather a single project manager. The effort and expense needed by the funding organization would be nearly identical between journal publishing and report publishing.
To protect academic freedom and ensure that a government was not overly influencing knowledge dispersal the operation of the journals (e.g. editorial decisions) would be governed by academics and authors would still have the right to additionally publish elsewhere. In addition, these journals would be open for submissions from any authors including those not funded by the specific organization operating the journal. The articles would undergo the same peer-review they do now in any conventional journal. The editors would similarly be drawn from the scientific community and the journal could be managed by existing project managers during time released from eliminating reporting requirements. The OA Journals would only be published digitally using the existing free Public Knowledge Project's Open Journal System, which already supports over 10,000 journals. The costs of type setting could be eliminated using either Libre Office or LaTeX templates. The final remaining cost associated with journal publishing is copy editing, which can be comfortably accomplished for less than a tenth of standard open access fees at major publishers.
This approach was targeted specifically at the U.S., but should work equally well in any country that has government funding of research in any discipline. In fact, the more countries that participate the faster the benefits of open access to a greater percentage of the literature accrue to the entire scientific community. For narrowly defined disciplines researchers would publish in the open access journal sponsored by their funders. However, as shared funding models and interdisciplinary work becomes more important in some areas of research rules would allow authors to publish in the most appropriate OA journals for the specific aspects of their research.
After a few years of all the top government-funded researchers publishing in sponsored OA journals, their impact factors would rise both from the volume and because of the open access effect (academics tend to cite articles more often if they can read them and unfortunately must of the literature is inaccessible to many academics). This proposal would be for totally “free to submit to” and “free to read” journals. In addition, these journals would have the seal of approval from the government funding agencies to avoid the challenges of independent OA journals, who often have their reputations attacked by traditional publishers and their surrogates.
Establishing such free OA journals would attract and encourage academics funded elsewhere to submit their papers to these high-quality open access journals for free (e.g. no article processing charge or APC). It should be noted, that APCs represent a loss of scientific funding. This would result in the vast majority of the literature being available without subscriptions or fees, which in turn would accelerate science in every field. As academics would be using their time more efficiently there would be more available research time allocated to do research, while the time overall spent on administrative tasks for academics would decrease. This will increase the rate of scientific discovery and the number of articles would increase. Thus academics would spend more time reviewing, but their overall administrative time would still be reduced. Elimination of the need for journal subscriptions in turn would lower overhead rates at universities, releasing more funds for actual research and improving funding rates rather than subsidizing stock holders of for-profit academic publishers. Everyone else including funders, researchers, the general public and mother science all win- big!